MicroRNA regulation in experimental autoimmune encephalomyelitis in mice and marmosets resembles regulation in human multiple sclerosis lesions

Journal of Neuroimmunology 246 (2012) 27–33

Contents lists available at SciVerse ScienceDirect

Journal of Neuroimmunology journal homepage: www.elsevier.com/locate/jneuroim

MicroRNA regulation in experimental autoimmune encephalomyelitis in mice and marmosets resembles regulation in human multiple sclerosis lesions Juliane Lescher a, Franziska Paap a, Verena Schultz a, Laura Redenbach a, Uta Scheidt a, Hendrik Rosewich b, Stefan Nessler a, Eberhard Fuchs c, Jutta Gärtner b, Wolfgang Brück a, Andreas Junker a,⁎ a b c

Department of Neuropathology, University of Göttingen, Göttingen, Germany Department of Pediatrics and Pediatric Neurology, University of Göttingen, Göttingen, Germany Clinical Neurobiology Laboratory, German Primate Center, Göttingen, Germany

a r t i c l e

i n f o

Article history: Received 5 February 2012 Accepted 22 February 2012 Keywords: MicroRNA Multiple sclerosis EAE miRNA-155 miRNA-146a/b Marmoset

a b s t r a c t Here we demonstrate that miRNA regulation in marmoset (Callithrix jacchus) and C57/BL6 mouse EAE lesions largely resembles miRNA regulation in active human MS lesions. Detailed quantitative PCR analyses of the most up- and downregulated miRNAs of active human MS lesions in dissected lesions from marmoset EAE brains and inflamed spinal cords of EAE mice revealed that the conserved and highly regulated miRNAs, miRNA-155, miRNA-142-3p, miRNA-146a, miRNA-146b and miRNA-21, turned out to be similarly upregulated in marmoset and mouse EAE lesions. © 2012 Elsevier B.V. All rights reserved.

1. Introduction MicroRNAs (miRNAs) are small (~22 bp) RNA fragments with regulatory functions on the translational level (Baek et al., 2008; Selbach et al., 2008). Recently, new information about miRNA expression in multiple sclerosis (MS) and its animal models was uncovered and functions of these miRNAs have been defined (for review see Junker, 2011; Junker et al., 2011; Thamilarasan et al., 2012). The expression of more than one-third of all mammalian genes might be regulated by miRNAs (Lewis et al., 2005) and they might therefore serve as therapeutic targets (Krutzfeldt et al., 2005; Czech, 2006; Elmen et al., 2008). It has become clear in recent years that human miRNAs are also highly conserved especially in other mammals (Lagos-Quintana et al., 2001; Lau et al., 2001; Lee and Ambros, 2001). It is increasingly recognized that miRNAs play essential roles in the immune system (Taganov et al., 2007; Lodish et al., 2008; O'Connell et al., 2010b) and for correct function in the mature central nervous system (CNS) (Kosik, 2006; Coolen and Bally-Cuif, 2009). It was found that miRNAs are critical for the maintenance of immune tolerance, as can be seen in mice with a deletion of Dicer-mediated miRNA processing in regulatory T cells (T reg cells). These mice develop a fatal autoimmune

⁎ Corresponding author at: Department of Neuropathology, University Medical Center, Georg August University Göttingen, Robert-Koch-Str. 40, 37075 Göttingen, Germany. Tel.: +49 551 398469; fax: +49 551 398472. E-mail address: [email protected] (A. Junker). 0165-5728/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.jneuroim.2012.02.012

disease which is very similar to the autoimmune syndrome of animals with a complete deficiency of T reg cells (Liston et al., 2008; Zhou et al., 2008). miRNA-124 promotes microglia quiescence and inhibits the symptoms of experimental autoimmune encephalomyelitis (EAE) by deactivating macrophages via the C/EBP-α-PU.1 pathway (Ponomarev et al., 2011). Apart from this, inflammatory cytokines can regulate the expression levels of miRNAs in a distinct cell type as shown, for example, in astrocytes (Junker et al., 2009; Mor et al., 2011). Only recently, several studies have been published in which expression of miRNAs was measured in the peripheral blood of MS patients (Du et al., 2009; Keller et al., 2009; Otaegui et al., 2009; Cox et al., 2010; De Santis et al., 2010; Lindberg et al., 2010; Martinelli-Boneschi et al., 2012; Fenoglio et al., 2011; Guerau-deArellano et al., 2011; Waschbisch et al., 2011). Two studies examined expression profiles of miRNAs in the CNS of MS patients (Junker et al., 2009; Noorbakhsh et al., 2011). miRNA-155 was one of the most upregulated miRNAs in active human MS lesions (Junker et al., 2009) and is also upregulated in normal appearing white matter (NAWM) from the CNS of MS patients (Noorbakhsh et al., 2011). Mice with a deficiency of miRNA-155 develop less severe EAE than controls (O'Connell et al., 2010a; Murugaiyan et al., 2011). For example, the attenuation of EAE in miRNA-155 knockout mice was associated with a decrease in Th1 and Th17 responses in the CNS and peripheral lymphoid organs. MS is an inflammatory disease of the CNS which leads to the formation of demyelinated plaques with glial scar formation and axonal loss (Noseworthy et al., 2000). The heterogeneity of MS lesions, including the inflammatory infiltrate, oligodendrocyte damage and

28

J. Lescher et al. / Journal of Neuroimmunology 246 (2012) 27–33

the extent of demyelination/remyelination or axonal loss, has recently been described (Lucchinetti et al., 1999; Kuhlmann et al., 2002). The detailed histopathological description of actively demyelinating MS lesions led to the definition of four distinct lesion patterns (Lucchinetti et al., 2000). Several animal models displaying parts of the pathology of MS and CNS inflammation have now been established (for review see Krishnamoorthy and Wekerle, 2009). EAE induced with the myelin oligodendrocyte glycoprotein peptide MOG35–55 in C57/BL6 mice is one common model for investigating the CNS inflammation. The striking features of this model are large acute inflammatory lesions within the spinal cord with a great number of infiltrating CD4 + Tcells and macrophages with demyelination. The inflammatory CNS lesions in mice with MOG35–55 EAE bear features that are comparable to human MS lesions. MOG1–125-induced EAE in the common marmoset (Callithrix jacchus) is a model of CNS inflammation which displays many parallels, with regard to the pathology of the lesions, when compared with active human MS lesions (Merkler et al., 2006b). We asked whether miRNA regulation in two inflammatory animal models, namely the MOG35–55 peptide-induced EAE in mice and the MOG1–125-induced EAE in marmosets, is comparable to the miRNA regulation of active human MS lesions. Here we show that the expression of the most upregulated miRNAs in MS largely resembles the miRNA expression in these two EAE models. 2. Material and methods 2.1. Tissue specimens of marmosets In total, we analyzed 16 formalin-fixed paraffin-embedded (FFPE) tissue samples from marmoset brains. The group of marmoset tissue specimens comprised 10 FFPE blocks with EAE lesions and 6 control blocks derived from animals which did not develop any lesion pathology after immunization (n = 5) or did not receive any immunization at all (n = 1). Tissue samples originated from the Department of Neuropathology in Göttingen, Germany and were collected from previous projects (Merkler et al., 2006a; Merkler et al., 2006b; Diem et al., 2008). Briefly, marmosets had been immunized with recombinant protein (rMOG) corresponding to the N-terminal sequence of rat MOG (amino acids 1–125) in complete Freund's adjuvant (CFA) (Merkler, D. et al., 2006b). All experiments had been performed in compliance with relevant laws and institutional guidelines and had been approved by local authorities. 2.2. Tissue specimens of mice MOG35–55 (MEVGWYRSPFSRVVHLYRNGK) was purchased from Princeton Biomolecules (Langhorne, PA, USA). Incomplete Freund's adjuvant (IFA) was prepared as a mixture of mannide monooleate (Sigma-Aldrich, Steinheim, Germany) and paraffin oil (Merck, Darmstadt, Germany); CFA was obtained by mixing Mycobacterium tuberculosis H37RA (Difco Laboratories, Franklin Lakes, NJ, USA) at 5 mg/ ml into IFA. For disease induction 20 C57/BL6 mice were immunized 128 s.c. in all four extremities with a total of 200 μg MOG35–55 in CFA. 300 ng pertussis toxin (PTX; List Biological Laboratories — Quadratech, Epsom, UK) in 300 μl sterile phosphate buffered saline (PBS) was given on the day of immunization and 48 h later. Ten nonimmunized mice served as negative controls. Animals were weighed and scored daily for clinical signs of disease on a scale from 0 to 10. Scores were as follows: 0 = normal; 1 = reduced tone of tail; 2 = limp tail, impaired righting; 3 = absent righting; 4 = gait ataxia; 5 = mild paraparesis of hind limbs; 6 = moderate paraparesis; 7 = severe paraparesis or paraplegia; 8 = tetraparesis; 9 = moribund; and 10 = death (Wust et al., 2008). Mice that either developed a score above 6 or had EAE symptoms for more than 4 days were sacrificed.

All experiments were performed in compliance with relevant laws and institutional guidelines and were approved by local authorities. 2.3. Macrodissection of marmoset lesions Lesions of marmoset brain samples were extracted as described previously for human brain samples (Eisele et al., 2012; Junker et al., 2009). Briefly, paraffin sections were mounted on membranecovered polyethylene naphthalate slides (Zeiss, Jena, Germany). Parallel sections were stained with hematoxylin–eosin (HE) and luxol fast blue-Pas (LFB-Pas) to allow identification of the lesions. For the analysis of miRNA profiles from lesions, white matter areas with inflammatory infiltrate were dissected from the slides with a scalpel and 10 sections (each 9 μm thick) were pooled for RNA extraction. Control white matter specimens were extracted the same way. This miRNA analysis was restricted to white matter tissue samples in order to limit possible confounding effects from neuronal miRNAs. FFPE tissue was deparaffinized and digested with proteinase K (Sigma-Aldrich) before RNA extraction. 2.4. Processing of spinal cord tissue of mice Spinal cords of diseased mice or control mice were extracted and snap frozen for histochemistry and RNA preparation. 2.5. Histopathology Histological evaluation was performed on 3 μm-thick sections stained with HE (data not shown) and LFB-Pas to assess inflammation and demyelination. Immunohistochemistry was performed after antigen-unmasking microwave treatment for 15 min (800 W) in citrate buffer. Endogenous peroxidase activity was blocked by incubation of sections in 3% H2O2 in PBS. Sections were blocked with 10% fetal calf serum in PBS for 10 min at room temperature. Washed sections were stained with the following primary antibodies: mouse anti-human MRP14 (1:500, BMA Biomedicals, Augst, Switzerland), mouse anti-human CD3 (1:50, Serotec, Düsseldorf, Germany), rat anti-mouse MAC3 — M3/84 (1:200, Pharmingen, Hamburg, Germany). Bound antibody was visualized using an avidin-biotin technique with 3,3′-diaminobenzidine (DAB) as chromogen. 2.6. RNA extraction Tissue lysis and RNA extraction were performed according to the protocol previously published by Eisele et al. and Junker at al. (Eisele et al., 2012; Junker et al., 2009). FFPE tissue and frozen tissue are equally applicable for the detection of miRNAs as demonstrated before (Eisele et al., 2012; Junker et al., 2009). For tissue lysis of FFPE specimens lysis buffer (50 mMTris, 25 mM EDTA, 500 mM NaCl, 0.1% Nonidet® P-40, 1% SDS) in combination with 1/10 volume Proteinase K (Sigma-Aldrich) was added per reaction tube and incubated at 60 °C for 18 h. To improve tissue lysis reaction tubes were vortexed every 30 min at 1000 rpm for 5 min. RNA was isolated from the preprocessed FFPE specimens or directly from the frozen specimens using the miRNeasy mini-kit (Qiagen, Hilden, Germany) according to the protocol provided by the manufacturer. To obtain a higher yield of RNA, we used RNeasy microspin columns (Qiagen) RNA concentration was determined with the NanoDrop (Thermo scientific/Gerhard Menzel, Braunschweig, Germany). RNA was stored at −80 °C. 2.7. RT PCR — Taqman LDAs Evaluation of miRNA profiles of mouse samples was performed as described previously for human miRNAs (Junker et al., 2009). Briefly, the investigated screening samples contained 4 EAE spinal cords and

J. Lescher et al. / Journal of Neuroimmunology 246 (2012) 27–33

3 control spinal cords from non-diseased mice. In total, the expression level of 586 single miRNAs was detected with LDAs on the ABI 7900 (Applied Biosystems, Darmstadt, Germany). For analyses of mouse data the median of the most abundant 228 miRNAs (228 miRNAs could be detected with a raw cycle threshold (CT b30) in all samples) was used as a surrogate housekeeping gene as previously described (Junker et al., 2009). The relative expression of miRNAs in EAE lesions versus control tissue was calculated with the ΔCT method (Ng et al., 2009). The qPCR single reactions were performed on the IQ5 (Biorad, Munich, Germany) using the qPCR Core Kit and uracyl N-glycosylase (both from Eurogentec, Cologne, Germany). miRNAs were detected with single TaqMan miRNA Assays (Applied Biosystems). The total amount of transcribed RNA equivalent used per PCR reaction was 1 ng for all miRNAs. In marmoset experiments, the small nuclear RNA RNU6B was used for normalization with the ΔCT method (Ng et al., 2009). Briefly, ΔCT = [(median of CT miRNA lesions minus median CT RNU6B lesions) minus (CT miRNA controls minus CT RNU6B controls)]. In mice, RNU6B appeared to be upregulated in mouse EAE lesions when normalized against the amount of RNA equivalent used per qPCR reaction (data not shown). Therefore, 4 miRNAs which appeared not to be 214 regulated in the mouse LDA screening experiment (miRNA-135b, miRNA-125a-5p, miRNA-132, miRNA-491), were measured in each sample and the median raw CT value of all 4 was used for normalization as a surrogate housekeeping gene for this experiment. The stability of the housekeeping gene or surrogate housekeeping gene in the examined samples is depicted in Supplementary Fig. 1. 2.8. Human samples Evaluation of miRNA profiles of human samples was previously published and described in detail (Junker et al., 2009). In short, the investigated samples contained 16 active MS lesions and 9 normal brain white matter specimens (for details see Junker et al., 2009). In short, the investigated samples contained 16 active MS lesions and 9 normal brain white matter specimens (for details see (Junker et al., 2009)).

29

2.9. Statistical analysis To assess whether two samples of independent observations varied in their composition, the Mann–Whitney U test was performed. The significance level was set to 5%. The correlation between miRNA regulation in human MS lesions and animal EAE lesions was calculated with Spearman's rank correlation. 3. Results 3.1. Histopathological analysis of the inflamed lesions The inflammatory infiltrates in demyelinating white matter EAE lesions from marmosets mainly comprise activated macrophages/ microglia (visualized with MRP14 antibody immunohistochemistry) and to a lesser extent CD3 + T-cells (Fig. 1). In marmosets, myelin stain with LFB-Pas revealed large confluent demyelinating lesions (Fig. 1). The inflammatory infiltration was partially perivascularly accentuated. The majority of the plaques in marmosets were early active and only small areas of the hemispheres contained inactive lesion areas. Detailed analyses of the lesions revealed a histopathological pattern which resembled the pathology of MS pattern II lesions (Lucchinetti et al., 2000; Merkler, D. et al., 2006b). Actively inflamed EAE lesions of mouse spinal cords displayed demyelination (stained with LFB-Pas) and accumulations of activated macrophages/microglia in meningeal and parenchymal localizations (stained with LFB-Pas and immunohistochemistry against Mac3 — Fig. 1). T cell infiltration (detected with antibodies against CD3) occurred in the parenchyma and perivascular area (Fig. 1). Histological analysis of human MS lesions showed actively demyelinating plaques with an accumulation of activated macrophages and microglia together with a sparse T-cellular infiltration, often accentuated around vessels (Junker et al., 2009). Infiltration by macrophages and T cells in the animal models resembled substantially active human MS lesions. Demyelination in marmoset EAE lesions is comparable to that in human MS lesions. In contrast, spinal cord EAE lesions in mice show less demyelination.

Fig. 1. Histology of analyzed lesions in animal models. Formalin-fixed paraffin-embedded (FFPE) (marmosets) or frozen (mice) tissue blocks containing active demyelinating lesions were stained in parallel for myelin and macrophages with luxol fast blue-Pas (LFB-Pas), CD3 for T cells and MRP14 (marmosets) or Mac3 (mice) for macrophage infiltration. Demyelination detected with LFB-Pas is depicted in (A) for marmosets and (D) for mice. T-cellular infiltration is visualized in (B) for marmosets and (E) for mice. Active lesions are densely populated with macrophages as shown in (C) for marmosets and (F) for mice. Paraffin sections were 4 μm thick, frozen sections were 7 μm thick.

30

J. Lescher et al. / Journal of Neuroimmunology 246 (2012) 27–33

3.2. Overexpressed miRNAs in human MS are partially upregulated in mouse and marmoset EAE lesions

ΔΔCT ((ΔCCT surrogate housekeeping gene - CT miRNA)active human lesions ΔΔ -(ΔCT surrogate housekeeping gene - CT miRNA)controls))

3.2.1. miRNA regulation in spinal cords of mice with MOG35–55-induced EAE We performed screening experiments with qPCR (Taqman miRNA LDAs — Taqman Array Rodent miRNA Panels A and B) to determine the expression level of 586 single miRNAs in EAE mouse spinal cords and spinal cords of non-diseased mice. In this screening experiment, only a limited number of specimens (4 EAE spinal cords and 3 control spinal cords from non-diseased mice) were investigated. We could detect 408 miRNAs in the mouse spinal cords. Thirty-eight miRNAs appeared to be upregulated more than two-fold in inflamed spinal cords compared to spinal cords of non-diseased mice, and 41 miRNAs were downregulated by less than a factor of 0.5 in inflamed mouse spinal cords versus non-inflamed mouse spinal cords (for complete dataset, see Supplementary Table 1). We considered miRNA abundance in human tissue and mice spinal cords (CT b35) and compared the significantly (p b 0.05) regulated miRNAs of active human MS lesions with corresponding miRNAs from the qPCR — LDAs of inflamed spinal cords. We found that the regulation of human miRNAs in MS lesions partially overlapped with this animal model (Fig. 2 and Supplementary Table 2). To confirm these results and to follow the hypothesis of partially convergent miRNA regulation in animal models and humans, we measured the expression of miRNAs with Taqman single miRNA assays (Applied Biosystems) in inflamed mouse spinal cords (17 specimens from different animals) and spinal cords from non-diseased mice (10 specimens from different animals). All conserved miRNAs which were either upregulated 3.5 fold (8 miRNAs) or downregulated by less than a factor of 0.29 (1 miRNA) in the active human white matter MS brain lesions and a selection of

unregulated miRNAs (5 miRNAs) – based on our previous report (Junker et al., 2009) – were measured in inflamed spinal cord sections of mice with MOG35–55 peptide-induced EAE (Table 1). Six of the conserved miRNAs were likewise significantly upregulated in active mouse EAE lesions compared to controls (miRNA-155, miRNA-326, miRNA-142-3p, miRNA-146a, miRNA-146b and miRNA-21) (Table 1 and Supplementary Fig. 2). One miRNA showed a 4.3-fold upregulation which did not reach a level of significance (miRNA-23a). miRNA-34a showed a tendency towards upregulation (1.5 fold) in mouse EAE lesions which was also non-significant. miRNA-184, the only miRNA which was downregulated in human MS lesions by less than 0.29 fold and which is also conserved in mice, showed – in clear contrast to the human disease – a 2.5-fold upregulation. Of the 5 selected miRNAs in human MS lesions which were not regulated compared to controls (no significant differences between lesions and control tissue — Table 1), only 4 showed comparable results in mice (miRNA-135b, miRNA125a-5p, miRNA-132 and miRNA-491). miRNA-19a, which was not regulated in human MS lesions, showed a 2.5-fold upregulation in mouse EAE spinal cords. Correlation between miRNA regulation in human MS lesions and miRNA regulation in mouse EAE lesions showed moderate correlation (Correlation coefficient r = 0.67; p b 0.01; Spearman's rank correlation). Absolute regulation differs in the qPCR single assay investigations as compared to the LDA screening analyses. The expression levels of the screening experiment have only an indicative character due to the low sample size and could therefore not reach significant level. 3.2.2. miRNA regulation in marmoset EAE lesions In marmoset tissue, 8 conserved miRNAs, which were upregulated more than 3.5 fold and 1 miRNA which was downregulated by less than 0.29 fold in active human MS lesions (Junker et al., 2009), were analyzed. Altogether we investigated 10 active white matter

5,00

4,00

3,00 R2 = 0,3888

2,00

1,00

-2,00

-1,00

0,00 0,00

1,00

2,00

3,00

4,00

5,00

-1,00

-2,00

-3,00 ΔΔCT ((ΔCT surrogate housekeeping gene - CT miRNA)mouse EAE lesions ΔΔ -(ΔCT surrogate housekeeping gene - CT miRNA)controls))

Fig. 2. miRNA expression in multiple sclerosis lesions in comparison to miRNA expression of mouse EAE lesions (screening experiment). Comparison of miRNAs up- or downregulated in active human lesions versus their corresponding miRNAs in mouse experimental autoimmune encephalomyelitis (EAE) spinal cords measured with Taqman Low Density Arrays (LDAs) (Applied Biosystems, Germany). The X-value of each dot represents the regulation of one specific miRNA in EAE lesions versus corresponding controls [ΔΔCT value = ((ΔCT surrogate housekeeping gene − CT miRNA) mouse EAE lesions − (ΔCT surrogate housekeeping gene − CT miRNA) controls)]. The Y-value of each dot represents the regulation of the same miRNA in active human MS lesions versus controls [ΔΔCT value=((ΔCT surrogate housekeeping gene−CT miRNA) MS lesions−(ΔCT surrogate housekeeping gene−CT miRNA) controls)]. Values are listed in Supplementary Table 2. Coefficient of determination R2 = 0.389.

J. Lescher et al. / Journal of Neuroimmunology 246 (2012) 27–33 Table 1 Regulation of selected miRNAs of active human multiple sclerosis lesions in comparison to mouse and marmoset EAE lesions. miRNAsa

miRNA-155 miRNA-326 miRNA-142-3p miRNA-146a miRNA-146b miRNA-34a miRNA-21 miRNA-23a miRNA-184 miRNA-135b miRNA-125a-5p miRNA-132 miRNA-491 miRNA-19a

Human

Mouse

Marmoset

Fold regulation in lesions compared to controlb

Fold regulation in lesions compared to controlc

Fold regulation in lesions compared to controld

11.9 8.9 7.7 6.3 5.0 4.9 3.9 3.9 0.2 0.9 1.0 0.7 1.0 1.2

⁎⁎ ⁎⁎ ⁎⁎ ⁎⁎ ⁎⁎ ⁎⁎ ⁎⁎ ⁎⁎ ⁎⁎ –e –e –e –e –e

17.0 1.7 26.8 3.0 2.9 1.5 10.8 4.3 2.5 0.9 1.2 1.0 1.2 2.5

⁎⁎ ⁎⁎ ⁎⁎ ⁎⁎ ⁎⁎ –e ⁎⁎ –e ⁎⁎ –e –e –e –e ⁎⁎

8.0 1.0 16.7 4.7 4.1 1.3 4.4 2.0 1.3 0.6 0.8 0.4 0.5 1.7

⁎⁎ –e ⁎⁎ ⁎⁎ ⁎⁎ –e ⁎⁎ –e –e –e –e ⁎⁎ ⁎⁎ ⁎⁎

a Selection of regulated (up- or downregulated at least 3.5 fold) and unregulated miRNAs of active human MS lesions in comparison to normal brain white matter (adopted from Junker et al., 2009). b 16 active human MS lesions and 9 control white matter specimens were examined (adopted from Junker et al., 2009). c 17 mouse EAE spinal cord sections and 10 control specimens were examined. d 10 marmoset EAE lesions and 6 control white matter specimens were examined. e No significant difference in miRNA regulation between active lesions and control tissue (U-test). ⁎⁎ p b 0.05 significant difference in miRNA regulation between active lesions and control tissue (U-test).

lesions from different animals and 6 white matter specimens from non-diseased animals. Five miRNAs were likewise significantly upregulated in active marmoset EAE lesions (miRNA-155, miRNA-142-3p, miRNA-146a, miRNA-146b, miRNA-21) (Table 1 and in detail in Supplementary Fig. 2). One showed a two-fold upregulation compared to controls which was not significant (miRNA-23a). miRNA-326 and miRNA-34a, in contrast to active human MS lesions, were not regulated in active marmoset EAE lesions. From the selected 5 miRNAs which showed no regulation in active human MS lesions, 3 showed a significant regulation in marmoset EAE lesions. miRNA-132 and miRNA491 were significantly downregulated and miRNA-19a showed a significant upregulation. Correlation between miRNA regulation in human MS lesions and miRNA regulation in marmoset EAE lesions showed moderate correlation (Correlation coefficient r = 0.68; p b 0.01; Spearman's rank correlation). Technical replicates were performed and showed comparable results (data not shown). The absolute regulatory level of each single miRNA differs among the investigated species. 4. Discussion We matched the miRNA regulation of the most upregulated miRNAs of human MS lesions with that in two different animal models for MS, namely the MOG35–55 peptide-induced EAE in C57/BL6 mice and the MOG1–125-induced EAE in marmoset monkeys. The majority of the most regulated miRNAs of human active MS lesions are regulated in a similar direction in the two animal models. The upregulation of miRNAs in active lesions compared to normal CNS tissue is caused, at least to some extent, by the cellular infiltration of T cells and macrophages. Another reason for the alteration in miRNA expression is presumably due to an upregulation of miRNAs in cells resident in the brain, such as astrocytes and microglia. Astrocytes are the most abundant glia cell type of the CNS and they play a key role in immune reactions (Farina et al., 2007). miRNA-155 and miRNA-146a/b are upregulated in active MS lesions of humans and

31

EAE lesions of mice and marmosets. Two of them, miRNA-155 and miRNA-146a, are upregulated in primary human astrocytes in vitro upon stimulation with interleukin-1β (IL-1β) and tumor necrosis factor-α (TNF-α) (Junker et al., 2009). After exposure to interferonγ (IFN-γ) and lipopolysaccharide (LPS) in vitro miRNA-155, miRNA146a and miRNA-146b are upregulated in primary astrocytes of mice (Mor et al., 2011). This could also be demonstrated for miRNA155 and miRNA-146a in primary astrocytes of marmosets (Mor et al., 2011). miRNA-155, miRNA-34a and miRNA-326 target the immunomodulatory surface molecule CD47, which is downregulated in active and inactive MS lesions and which might enhance macrophage activity (Junker et al., 2009). This demonstrates that different upregulated miRNAs may share common target transcripts. To date, several studies have shown the proinflammatory role of miRNA-155 on diverse cell types. One of the properties of miRNA-155 is the promotion of T cell-dependent tissue inflammation. Mice with a deficiency of this miRNA are resistant to the development of active EAE (O'Connell, R. M. et al., 2010a). miRNA-155 was identified as a component of macrophage/monocyte response to LPS, IFN-β, TNF-α and polyriboinosinic-polyribocytidylic acid (poly IC) (O'Connell et al., 2007; Tili et al., 2007). Moreover, miRNA-155 modulates the immune response in microglia by downregulating suppressor of cytokine signaling 1 (SOCS-1) and increasing cytokine and NO production (Cardoso et al., 2011). This is of particular interest because much of the inflammatory infiltrate in the lesions consists of activated microglia and macrophages, and the activity of these cells seems to be crucial for the degree of devastation in the lesions (Benveniste, 1997; Sriram and Rodriguez, 1997; Platten and Steinman, 2005; Breij et al., 2008). Interestingly, miRNA-155 is also upregulated in NAWM of MS patients (Noorbakhsh et al., 2011). This is accompanied by suppressed expression of critical neurosteroidogenic enzymes (3α-hydroxysteroid dehydrogenase isoforms), which are targeted by miRNA-338, miRNA-155 and miRNA-491 (Noorbakhsh et al., 2011). In that study, several miRNAs were also checked for their expression in EAE. miRNA-155 as well as miRNA-338-3p showed upregulation in EAE hindbrains compared with controls. miRNA-338-5p was not altered in EAE, whereas analysis of miRNA-183 and miRNA-491 revealed a non-significant increase in EAE compared with controls (Noorbakhsh et al., 2011). This is in line with our results, in which miRNA-155 shows a strong upregulation in EAE spinal cords and miRNA-491 does not show a significant regulation in EAE versus controls. MiRNA-146a/b is involved in the regulation of the adaptive as well as the innate immune response (reviewed by Rusca and Monticelli, 2011) and might therefore play fundamental roles in astrocytes under inflammatory conditions. miRNA-146a was upregulated in reactive astrocytes from human hippocampal epilepsy biopsies and also in the hippocampus of a rat model of epilepsy (Aronica et al., 2010). Furthermore, this miRNA was increased in human Alzheimer's disease brain samples from the neocortex and hippocampus and in vitro in astrocytes treated with IL-1β or amyloid-β-42 (Aβ42) peptide (Cui et al., 2010). It was shown that Toll-like receptor and cytokine signaling are influenced by miRNA-146a, which targets the TNF receptor-associated factor 6 (TRAF6) and IL-1 receptor-associated kinase 1 (IRAK1) genes in primary human astrocytes (Cui et al., 2010); this was previously shown in monocytes (Taganov et al., 2006). NFkappaB-sensitive miRNA-146a-mediated modulation of complement factor H gene expression may in part regulate an inflammatory response in Alzheimer's disease brain and in stressed human glial cells in vitro (Lukiw et al., 2008). In whole tissue miRNA analyses of MS lesions as well as of EAE lesions, miRNA-146a probably reflects the miRNA content of both parenchymal and infiltrating cells, where it also covers important regulatory functions. miRNA-146a is upregulated in Th1 and downregulated in Th2 T cells in mice and might contribute to determine the fate of these cells (Monticelli et

32

J. Lescher et al. / Journal of Neuroimmunology 246 (2012) 27–33

al., 2005). Furthermore, miRNA-146a is important for the function of T reg cells and a deficiency of this miRNA results in impaired T reg function with the breakdown of immunological tolerance, massive lymphocyte activation and tissue infiltration in several organs (Lu et al., 2010). In our study we demonstrate that miRNA-326, upregulated in active human MS lesions (Junker et al., 2009), is also upregulated in mouse EAE lesions — but not in marmoset EAE lesions. Besides its potential regulatory function on CD47 in MS lesions (Junker et al., 2009), miRNA-326 expression in peripheral blood is correlated with disease severity in MS patients and EAE mice. Inhibition of miRNA-326 was shown to reduce EAE symptoms in mice by inhibiting Th17 cell differentiation through a T cell-intrinsic mechanism. It was shown that miRNA-326 targets Ets-1, a known negative regulator of Th17 cells (Du et al., 2009). miRNA-326, miRNA-155, miRNA-146a and miRNA142-3p were overexpressed in peripheral blood mononuclear cells (PBMC) derived from patients with relapsing-remitting MS (RR-MS) and were expressed partially lower in glatiramer acetate (GA)-treated RR-MS patients (Waschbisch et al., 2011). Upregulation of miRNA-326 might be correlated to inflammatory activity in the CNS, because of its stronger upregulation in the brains of a subset of MS patients with the so-called Marburg variant of the disease (Junker et al., 2009). Marburg variant is a very aggressive form of MS with large and strongly inflamed brain lesions and a poor clinical prognosis (Hu and Lucchinetti, 2009). The heterogeneity of the inflammatory infiltrate might also be the reason why miRNA-34a is not significantly regulated in the animal lesions as opposed to active human MS lesions. miRNA-34a is differentially expressed in human T cell subsets from patients with RR-MS (Lindberg et al., 2010) and the T-cellular infiltration differs in its composition between human MS lesions and both of the animal EAE models (Table 1). In humans, extended parts of the T-cellular infiltration of the lesions consist of CD8+ T-cells, which frequently outrange CD4+ T cells in number and which have a key role in the pathology of MS (reviewed by Saxena et al., 2011). Furthermore, grey matter in mouse spinal cords might contribute to a shift in miRNA expression, since only white matter lesions were analyzed in human MS (Junker et al., 2009) and marmoset specimens. Another factor which might have influence on lesion pathology is the up- or downregulation of non-conserved miRNAs in human MS lesions compared to the animal models. miRNA-650 is the most upregulated miRNA in active human MS lesions (Junker et al., 2009) and it is not conserved in mice or marmosets; the same is true for the non-conserved miRNA-656 which is downregulated in active human MS lesions (Junker et al., 2009) and does not occur in mice or marmosets (miRNA gene identifying algorithms see miRviewer http://people.csail.mit.edu/akiezun/ miRviewer/). Targets of these non-conserved miRNAs might therefore not be affected in the animal models of MS. Non-conserved binding sites might also lead to differences in the effects of miRNA in human MS or the animal models. miRNA-298 and miRNA-351, for example, are downregulated in mouse astrocytes and are specific to mice and rats (Mor et al., 2011). Thus the TNF-α 3′ UTR binding site for miRNA-298 is conserved among most species, while the miRNA-351/miRNA-125b binding site is found in mice but not in primates (Mor et al., 2011). The same is true for other factors, such as the E3 ubiquitin-protein ligase seven in absentia homolog 2 (SIAH2) (Mor et al., 2011), which mediates TNF-α-induced ubiquitination and degradation of TRAF2 (Habelhah et al., 2002). To summarize, in this study we investigated miRNA expression of the most up- or downregulated miRNAs of active human MS lesions in two inflammatory animal models, the MOG35–55 peptide-induced EAE in C57/BL6 mice and the MOG1–125-induced EAE in the common marmoset monkey. We show that the conserved and most upregulated miRNAs of active human MS lesions, miRNA-155, miRNA-1423p, miRNA-146a/b and miRNA-21, are also upregulated in active

lesions in the preclinical models. This underlines that basic immune mechanisms modulated by miRNAs in active human MS lesions are comparable to the ones in these two inflammatory disease models. Acknowledgments We are grateful to Nancy Meyer for useful discussions and help with the scoring of EAE mice. The authors thank Christine Crozier for critical reading of the manuscript. Appendix A. Supplementary data Supplementary data to this article can be found online at doi:10. 1016/j.jneuroim.2012.02.012. References Aronica, E., Fluiter, K., Iyer, A., et al., 2010. Expression pattern of miR-146a, an inflammation-associated microRNA, in experimental and human temporal lobe epilepsy. Eur. J. Neurosci. 31, 1100–1107. Baek, D., Villen, J., Shin, C., Camargo, F.D., Gygi, S.P., Bartel, D.P., 2008. The impact of microRNAs on protein output. Nature 455, 64–71. Benveniste, E.N., 1997. Role of macrophages/microglia in multiple sclerosis and experimental allergic encephalomyelitis. J. Mol. Med. (Berl.) 75, 165–173. Breij, E.C., Brink, B.P., Veerhuis, R., et al., 2008. Homogeneity of active demyelinating lesions in established multiple sclerosis. Ann. Neurol. 63, 16–25. Cardoso, A.L., Guedes, J.R., de Almeida, L.P., Pedroso de Lima, M.C., 2011. miR-155 modulates microglia-mediated immune response by downregulating SOCS-1 and promoting cytokine and NO production. Immunology 10–2567. Coolen, M., Bally-Cuif, L., 2009. MicroRNAs in brain development and physiology. Curr. Opin. Neurobiol. 19, 461–470. Cox, M.B., Cairns, M.J., Gandhi, K.S., et al., 2010. MicroRNAs miR-17 and miR-20a inhibit T cell activation genes and are under-expressed in MS whole blood. PLoS One 5, e12132. Cui, J.G., Li, Y.Y., Zhao, Y., Bhattacharjee, S., Lukiw, W.J., 2010. Differential regulation of interleukin-1 receptor-associated kinase-1 (IRAK-1) and IRAK-2 by microRNA146a and NF-kappaB in stressed human astroglial cells and in Alzheimer disease. J. Biol. Chem. 285, 38951–38960. Czech, M.P., 2006. MicroRNAs as therapeutic targets. N. Engl. J. Med. 354, 1194–1195. De Santis, G., Ferracin, M., Biondani, A., et al., 2010. Altered miRNA expression in T regulatory cells in course of multiple sclerosis. J. Neuroimmunol. 226, 165–171. Diem, R., Demmer, I., Boretius, S., et al., 2008. Autoimmune optic neuritis in the common marmoset monkey: comparison of visual evoked potentials with MRI and histopathology. Investig. Ophthalmol. Vis. Sci. 49, 3707–3714. Du, C., Liu, C., Kang, J., et al., 2009. MicroRNA miR-326 regulates TH-17 differentiation and is associated with the pathogenesis of multiple sclerosis. Nat. Immunol. 10, 1252–1259. Eisele, S., Krumbholz, M., Fischer, M.T., et al., 2012. Prospects of transcript profiling for mRNAs and microRNAs using formalin-fixed and paraffin-embedded dissected autoptic multiple sclerosis lesions. Brain Pathol. Epub ahead of print, 10–3639. Elmen, J., Lindow, M., Schutz, S., et al., 2008. LNA-mediated microRNA silencing in nonhuman primates. Nature 452, 896–899. Farina, C., Aloisi, F., Meinl, E., 2007. Astrocytes are active players in cerebral innate immunity. Trends Immunol. 28, 138–145. Fenoglio, C., Cantoni, C., De, R.M., et al., 2011. Expression and genetic analysis of miRNAs involved in CD4+ cell activation in patients with multiple sclerosis. Neurosci. Lett. 504, 9–12. Guerau-de-Arellano, M., Smith, K.M., Godlewski, J., et al., 2011. Micro-RNA dysregulation in multiple sclerosis favours pro-inflammatory T-cell-mediated autoimmunity. Brain. 134, 3578–3589. Habelhah, H., Frew, I.J., Laine, A., et al., 2002. Stress-induced decrease in TRAF2 stability is mediated by Siah2. EMBO J. 21, 5756–5765. Hu, W., Lucchinetti, C.F., 2009. The pathological spectrum of CNS inflammatory demyelinating diseases. Semin. Immunopathol. 31, 439–453. Junker, A., 2011. Pathophysiology of translational regulation by microRNAs in multiple sclerosis. FEBS Lett. 585, 3738–3746. Junker, A., Krumbholz, M., Eisele, S., et al., 2009. MicroRNA profiling of multiple sclerosis lesions identifies modulators of the regulatory protein CD47. Brain 132, 3342–3352. Junker, A., Hohlfeld, R., Meinl, E., 2011. The emerging role of microRNAs in multiple sclerosis. Nat. Rev. Neurol. 7, 56–59. Keller, A., Leidinger, P., Lange, J., et al., 2009. Multiple sclerosis: microRNA expression profiles accurately differentiate patients with relapsing-remitting disease from healthy controls. PLoS One 4, e7440. Kosik, K.S., 2006. The neuronal microRNA system. Nat. Rev. Neurosci. 7, 911–920. Krishnamoorthy, G., Wekerle, H., 2009. EAE: an immunologist's magic eye. Eur. J. Immunol. 39, 2031–2035. Krutzfeldt, J., Rajewsky, N., Braich, R., et al., 2005. Silencing of microRNAs in vivo with ‘antagomirs’. Nature 438, 685–689.

J. Lescher et al. / Journal of Neuroimmunology 246 (2012) 27–33 Kuhlmann, T., Lingfeld, G., Bitsch, A., Schuchardt, J., Bruck, W., 2002. Acute axonal damage in multiple sclerosis is most extensive in early disease stages and decreases over time. Brain 125, 2202–2212. Lagos-Quintana, M., Rauhut, R., Lendeckel, W., Tuschl, T., 2001. Identification of novel genes coding for small expressed RNAs. Science 294, 853–858. Lau, N.C., Lim, L.P., Weinstein, E.G., Bartel, D.P., 2001. An abundant class of tiny RNAs with probable regulatory roles in Caenorhabditis elegans. Science 294, 858–862. Lee, R.C., Ambros, V., 2001. An extensive class of small RNAs in Caenorhabditis elegans. Science 294, 862–864. Lewis, B.P., Burge, C.B., Bartel, D.P., 2005. Conserved seed pairing, often flanked by adenosines, indicates that thousands of human genes are microRNA targets. Cell 120, 15–20. Lindberg, R.L., Hoffmann, F., Mehling, M., Kuhle, J., Kappos, L., 2010. Altered expression of miR-17-5p in CD4+ lymphocytes of relapsing-remitting multiple sclerosis patients. Eur. J. Immunol. 40, 888–898. Liston, A., Lu, L.F., O'Carroll, D., Tarakhovsky, A., Rudensky, A.Y., 2008. Dicer-dependent microRNA pathway safeguards regulatory T cell function. J. Exp. Med. 205, 1993–2004. Lodish, H.F., Zhou, B., Liu, G., Chen, C.Z., 2008. Micromanagement of the immune system by microRNAs. Nat. Rev. Immunol. 8, 120–130. Lu, L.F., Boldin, M.P., Chaudhry, A., et al., 2010. Function of miR-146a in controlling Treg cell-mediated regulation of Th1 responses. Cell 142, 914–929. Lucchinetti, C., Bruck, W., Parisi, J., Scheithauer, B., Rodriguez, M., Lassmann, H., 1999. A quantitative analysis of oligodendrocytes in multiple sclerosis lesions. A study of 113 cases. Brain 122, 2279–2295. Lucchinetti, C., Bruck, W., Parisi, J., Scheithauer, B., Rodriguez, M., Lassmann, H., 2000. Heterogeneity of multiple sclerosis lesions: implications for the pathogenesis of demyelination. Ann. Neurol. 47, 707–717. Lukiw, W.J., Zhao, Y., Cui, J.G., 2008. An NF-kappaB-sensitive micro RNA-146amediated inflammatory circuit in Alzheimer disease and in stressed human brain cells. J. Biol. Chem. 283, 31315–31322. Martinelli-Boneschi, F., Fenoglio, C., Brambilla, P., et al., 2012. MicroRNA and mRNA expression profile screening in multiple sclerosis patients to unravel novel pathogenic steps and identify potential biomarkers. Neurosci. Lett. 508, 4–8. Merkler, D., Boscke, R., Schmelting, B., et al., 2006a. Differential macrophage/microglia activation in neocortical EAE lesions in the marmoset monkey. Brain Pathol. 16, 117–123. Merkler, D., Schmelting, B., Czeh, B., Fuchs, E., Stadelmann, C., Bruck, W., 2006b. Myelin oligodendrocyte glycoprotein-induced experimental autoimmune encephalomyelitis in the common marmoset reflects the immunopathology of pattern II multiple sclerosis lesions. Mult. Scler. 12, 369–374. Monticelli, S., Ansel, K.M., Xiao, C., et al., 2005. MicroRNA profiling of the murine hematopoietic system. Genome Biol. 6, R71. Mor, E., Cabilly, Y., Goldshmit, Y., et al., 2011. Species-specific microRNA roles elucidated following astrocyte activation. Nucleic Acids Res. 39, 3710–3723. Murugaiyan, G., Beynon, V., Mittal, A., Joller, N., Weiner, H.L., 2011. Silencing microRNA-155 ameliorates experimental autoimmune encephalomyelitis. J. Immunol. 187, 2213–2221. Ng, E.K., Chong, W.W., Jin, H., et al., 2009. Differential expression of microRNAs in plasma of patients with colorectal cancer: a potential marker for colorectal cancer screening. Gut 58, 1375–1381.

33

Noorbakhsh, F., Ellestad, K.K., Maingat, F., et al., 2011. Impaired neurosteroid synthesis in multiple sclerosis. Brain 134, 2703–2721. Noseworthy, J.H., Lucchinetti, C., Rodriguez, M., Weinshenker, B.G., 2000. Multiple sclerosis. N. Engl. J. Med. 343, 938–952. O'Connell, R.M., Taganov, K.D., Boldin, M.P., Cheng, G., Baltimore, D., 2007. MicroRNA155 is induced during the macrophage inflammatory response. Proc. Natl. Acad. Sci. U. S. A. 104, 1604–1609. O'Connell, R.M., Kahn, D., Gibson, W.S., et al., 2010a. MicroRNA-155 promotes autoimmune inflammation by enhancing inflammatory T cell development. Immunity 33, 607–619. O'Connell, R.M., Rao, D.S., Chaudhuri, A.A., Baltimore, D., 2010b. Physiological and pathological roles for microRNAs in the immune system. Nat. Rev. Immunol. 10, 111–122. Otaegui, D., Baranzini, S.E., Armananzas, R., et al., 2009. Differential micro RNA expression in PBMC from multiple sclerosis patients. PLoS One 4, e6309. Platten, M., Steinman, L., 2005. Multiple sclerosis: trapped in deadly glue. Nat. Med. 11, 252–253. Ponomarev, E.D., Veremeyko, T., Barteneva, N., Krichevsky, A.M., Weiner, H.L., 2011. MicroRNA-124 promotes microglia quiescence and suppresses EAE by deactivating macrophages via the C/EBP-alpha-PU.1 pathway. Nat. Med. 17, 64–70. Rusca, N., Monticelli, S., 2011. MiR-146a in immunity and disease. Mol. Biol. Int. 437301 Epub; 2011 Apr 7., 437301. Saxena, A., Martin-Blondel, G., Mars, L.T., Liblau, R.S., 2011. Role of CD8 T cell subsets in the pathogenesis of multiple sclerosis. FEBS Lett. 585, 3758–3763. Selbach, M., Schwanhausser, B., Thierfelder, N., Fang, Z., Khanin, R., Rajewsky, N., 2008. Widespread changes in protein synthesis induced by microRNAs. Nature 455, 58–63. Sriram, S., Rodriguez, M., 1997. Indictment of the microglia as the villain in multiple sclerosis. Neurology 48, 464–470. Taganov, K.D., Boldin, M.P., Chang, K.J., Baltimore, D., 2006. NF-kappaB-dependent induction of microRNA miR-146, an inhibitor targeted to signaling proteins of innate immune responses. Proc. Natl. Acad. Sci. U. S. A. 103, 12481–12486. Taganov, K.D., Boldin, M.P., Baltimore, D., 2007. MicroRNAs and immunity: tiny players in a big field. Immunity 26, 133–137. Thamilarasan, M., Koczan, D., Hecker, M., Paap, B., Zettl, U.K., 2012. MicroRNAs in multiple sclerosis and experimental autoimmune encephalomyelitis. Autoimmun. Rev. 11, 174–179. Tili, E., Michaille, J.J., Cimino, A., et al., 2007. Modulation of miR-155 and miR-125b levels following lipopolysaccharide/TNF-alpha stimulation and their possible roles in regulating the response to endotoxin shock. J. Immunol. 179, 5082–5089. Waschbisch, A., Atiya, M., Linker, R.A., Potapov, S., Schwab, S., Derfuss, T., 2011. Glatiramer acetate treatment normalizes deregulated microRNA expression in relapsing remitting multiple sclerosis. PLoS One 6, e24604. Wust, S., van den Brandt, J., Tischner, D., et al., 2008. Peripheral T cells are the therapeutic targets of glucocorticoids in experimental autoimmune encephalomyelitis. J. Immunol. 180, 8434–8443. Zhou, X., Jeker, L.T., Fife, B.T., et al., 2008. Selective miRNA disruption in T reg cells leads to uncontrolled autoimmunity. J. Exp. Med. 205, 1983–1991.